0bps

Strategy vault / Breakout

BEST Engulfing Breakout Strategy

Hello traders This is a simple algorithm for a Tradingview strategy tracking a convergence of 2 unrelated indicators. Convergence is the solution to my trading problems. It's a puzzle with infinite possibilities and…

Author
ChaoZhang
Complexity
Intermediate · 207 lines
Parameters
11
Backtest bar
30m · intraday
Tested on
BTC/USDT · futures

Hello traders

This is a simple algorithm for a Tradingview strategy tracking a convergence of 2 unrelated indicators.

Convergence is the solution to my trading problems. It's a puzzle with infinite possibilities and only a few working combinations.

Here's one that I like

  • Engulfing pattern
  • Price vs Moving average for detecting a breakout

Definition

Take out the notebooks :) and some coffee (good for focus). I'm bullish in coffee

The engulfing pattern is a two-candle reversal pattern. The second candle completely 'engulfs' the real body of the first one, without regard to the length of the tail shadows.

The bullish Engulfing pattern appears in a downtrend and is a combination of one red candle followed by a larger green candle The bearish Engulfing pattern appears in a downtrend and is a combination of one green candle followed by a larger red candle

Example: https://imgur.com/a/krDDUz4

We're bored sir... what's the point of all this?

In summary, an engulfing is a pattern to track reversals. (the whole TradingView audience stands up now giving a standing ovation) Adding the Price vs Moving average filters allows to track reversals with momentums (half of the audience collapsed because this is too awesome)

Ok sir... you picked up my interest

I included some cool backtest filters:

  • date range filtering
  • flexible take profit in USD value (plotted in blue)
  • flexible stop loss in USD value (plotted in red)

All the best Dave

backtest

IMG

Parameters

NameDefaultWhat it controls
v_input_1_close0Source Price vs MA: close | high | low | open | hl2 | hlc3 | hlcc4 | ohlc4
v_input_20Type of MA: SMA | RMA | EMA | WMA | VWMA | SMMA | KMA | TMA | HullMA | DEMA | TEMA
v_input_332MA Length
v_input_42017Backtest Start Year
v_input_5trueBacktest Start Month
v_input_6trueBacktest Start Day
v_input_72020Backtest Stop Year
v_input_812Backtest Stop Month
v_input_931Backtest Stop Day
v_input_102000Backtest Profit Goal (in USD)
v_input_11200Backtest STOP Goal (in USD)

Source Pine Script

/*backtest
start: 2022-04-24 00:00:00
end: 2022-05-23 23:59:00
period: 30m
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

//@version=4
//@author=Daveatt

StrategyName = "BEST Engulfing + MA"
ShortStrategyName = "BEST Engulfing + MA"

strategy(title=StrategyName, shorttitle=ShortStrategyName, overlay=true, 
 pyramiding=2, default_qty_value=500, precision=7, currency=currency.USD,
 commission_value=0.2,commission_type=strategy.commission.percent, initial_capital=10000)

includeEngulfing = true

includeMA = true
source_ma = input(title="Source Price vs MA", type=input.source, defval=close)
typeofMA = input(title="Type of MA", defval="SMA", options=["RMA", "SMA", "EMA", "WMA", "VWMA", "SMMA", "KMA", "TMA", "HullMA", "DEMA", "TEMA"])
length_ma = input(32, title = "MA Length", type=input.integer)

// ---------- Candle components and states
GreenCandle = close > open
RedCandle = close < open
NoBody = close==open
Body = abs(close-open)


// bullish conditions
isBullishEngulfing1 = max(close[1],open[1]) < max(close,open) and min(close[1],open[1]) > min(close,open) and Body > Body[1] and GreenCandle and RedCandle[1]
isBullishEngulfing2 = max(close[1],open[1]) < max(close,open) and min(close[1],open[1]) <= min(close,open) and Body > Body[1] and GreenCandle and RedCandle[1]

// bearish conditions
isBearishEngulfing1 = max(close[1],open[1]) < max(close,open) and min(close[1],open[1]) > min(close,open) and Body > Body[1] and RedCandle and GreenCandle[1]
isBearishEngulfing2 = max(close[1],open[1]) >= max(close,open) and min(close[1],open[1]) > min(close,open) and Body > Body[1] and RedCandle and GreenCandle[1]

// consolidation of conditions
isBullishEngulfing = isBullishEngulfing1 or isBullishEngulfing2
isBearishEngulfing = isBearishEngulfing1 or isBearishEngulfing2

//isBullishEngulfing = max(close[1],open[1]) < max(close,open) and min(close[1],open[1]) > min(close,open) and Body > Body[1] and GreenCandle and RedCandle[1]
//isBearishEngulfing = max(close[1],open[1]) < max(close,open) and min(close[1],open[1]) > min(close,open) and Body > Body[1] and RedCandle and GreenCandle[1]

Engulf_curr = 0 - barssince(isBearishEngulfing) + barssince(isBullishEngulfing)
Engulf_Buy = Engulf_curr < 0 ? 1 : 0
Engulf_Sell = Engulf_curr > 0 ? 1 : 0


// Price vs MM


smma(src, len) =>
    smma = 0.0
    smma := na(smma[1]) ? sma(src, len) : (smma[1] * (len - 1) + src) / len
    smma

ma(smoothing, src, length) => 
    if smoothing == "RMA"
        rma(src, length)
    else
        if smoothing == "SMA"
            sma(src, length)
        else 
            if smoothing == "EMA"
                ema(src, length)
            else 
                if smoothing == "WMA"
                    wma(src, length)
				else
					if smoothing == "VWMA"
						vwma(src, length)
					else
						if smoothing == "SMMA"
						    smma(src, length)
						else
							if smoothing == "HullMA"
								wma(2 * wma(src, length / 2) - wma(src, length), round(sqrt(length)))
							else
								if smoothing == "LSMA"
									src
								else
								    if smoothing == "KMA"
								        xPrice = src
                                        xvnoise = abs(xPrice - xPrice[1])
                                        nfastend = 0.666
                                        nslowend = 0.0645
                                        nsignal = abs(xPrice - xPrice[length])
                                        nnoise = sum(xvnoise, length)
                                        nefratio = iff(nnoise != 0, nsignal / nnoise, 0)
                                        nsmooth = pow(nefratio * (nfastend - nslowend) + nslowend, 2) 
                                        nAMA = 0.0
                                        nAMA := nz(nAMA[1]) + nsmooth * (xPrice - nz(nAMA[1]))
                                        nAMA
								    else
								        if smoothing == "TMA"
									        sma(sma(close, length), length)
						                else
							                if smoothing == "DEMA"
							                    2 * src - ema(src, length)
							                else
							                    if smoothing == "TEMA"
							                        3 * (src - ema(src, length)) + ema(ema(src, length), length) 
							                    else
		    							            src
		    							                

MA = ma(typeofMA, source_ma, length_ma)

plot(MA, color=#006400FF, title="MA breakout", linewidth=3)

macrossover  = crossover (source_ma, MA)
macrossunder = crossunder(source_ma, MA)

since_ma_buy = barssince(macrossover)
since_ma_sell = barssince(macrossunder)
macross_curr = 0 - since_ma_sell + since_ma_buy
bullish_MA_cond = macross_curr < 0 ?  1 : 0
bearish_MA_cond = macross_curr > 0 ? 1  : 0

posUp = (Engulf_Buy ? 1 : 0) + (bullish_MA_cond ? 1 : 0) 
posDn = (Engulf_Sell ? 1 : 0) + (bearish_MA_cond ? 1 : 0) 

conditionUP = posUp == 2 and posUp[1] < 2
conditionDN = posDn == 2 and posDn[1] < 2


sinceUP = barssince(conditionUP)
sinceDN = barssince(conditionDN)

// primary-first signal of the trend
nUP = crossunder(sinceUP,sinceDN)
nDN = crossover(sinceUP,sinceDN)


// and the following secondary signals

// save of the primary signal
sinceNUP = barssince(nUP)
sinceNDN = barssince(nDN)

buy_trend   = sinceNDN > sinceNUP
sell_trend  = sinceNDN < sinceNUP

// engulfing by
barcolor(nUP ? color.orange : na, title="Bullish condition")
barcolor(nDN ? color.yellow : na, title="Bearish condition")

isLong  = nUP
isShort = nDN

long_entry_price    = valuewhen(nUP, close, 0)
short_entry_price   = valuewhen(nDN, close, 0)

longClose   = close[1] < MA
shortClose  = close[1] > MA

///////////////////////////////////////////////
//* Backtesting Period Selector | Component *//
///////////////////////////////////////////////


StartYear = input(2017, "Backtest Start Year",minval=1980)
StartMonth = input(1, "Backtest Start Month",minval=1,maxval=12)
StartDay = input(1, "Backtest Start Day",minval=1,maxval=31)
testPeriodStart = timestamp(StartYear,StartMonth,StartDay,0,0)

StopYear = input(2020, "Backtest Stop Year",minval=1980)
StopMonth = input(12, "Backtest Stop Month",minval=1,maxval=12)
StopDay = input(31, "Backtest Stop Day",minval=1,maxval=31)
testPeriodStop = timestamp(StopYear,StopMonth,StopDay,0,0)

testPeriod() => true


//////////////////////////
//* Profit Component *//
//////////////////////////

input_tp_pips = input(2000, "Backtest Profit Goal (in USD)",minval=0)
input_sl_pips = input(200, "Backtest STOP Goal (in USD)",minval=0)


tp = buy_trend? long_entry_price + input_tp_pips : short_entry_price - input_tp_pips
sl = buy_trend? long_entry_price - input_sl_pips : short_entry_price + input_sl_pips


long_TP_exit  = buy_trend and high >= tp
short_TP_exit = sell_trend and low <= tp

plot(tp, title="TP", style=plot.style_circles, linewidth=3, color=color.blue)
plot(sl, title="SL", style=plot.style_circles, linewidth=3, color=color.red)

if testPeriod()
    strategy.entry("Long", 1, when=isLong)
    strategy.close("Long", when=longClose )
    strategy.exit("XL","Long", limit=tp,  when=buy_trend, stop=sl)


if testPeriod()
    strategy.entry("Short", 0,  when=isShort)
    strategy.close("Short", when=shortClose )
    strategy.exit("XS","Short", when=sell_trend, limit=tp, stop=sl)

Published by ChaoZhang on the FMZ strategy square · last modified 2022-05-25. Collected in the Quant Trading Vault. Unvalidated research: re-test with your own data and costs before trusting it.

Build this with AgenKit

A spec is not a trading system. Paste this into your coding agent and AgenKit runs a specialist team — architecture, test-first build, risk gates, review — with paper trading on by default.

in Claude Code, Cursor, Codex & 4 more
/agenkit build a production-grade trading bot with a backtest engine, risk controls and paper trading by default from strategies/BEST-Engulfing-Breakout-Strategy.md
  1. npx agenkit install engineering-kit in your project
  2. Paste the command into your agent and approve each phase
  3. Get tests, backtest, risk controls and a paper-trading switch

Understand it first

Before you run this, read the lesson behind the idea and how to backtest it honestly. Specs like this one are the survivors of an unknown number of trials; deflate their results accordingly.

Related specs

esc

Try , , or .

↑↓ move↵ open